This collaborative R01 application, entitled Improving Metabolic Parameters of Antipsychotic Child Treatment (IMPACT), is re-submitted in response to NIMH PA-07-092 (Collaborative R01s for Clinical and Services Studies of Mental Disorders, AIDS and Alcohol Use Disorders) and PA-07-078 (Treatment-Emergent Adverse Effects of Psychotropic Medication). This five-year, three-site study will be conducted at Johns Hopkins University/University of Maryland (M. Riddle PI), University of North Carolina (L. Sikich PI) and Zucker Hillside Hospital (C. Correll PI). The overarching goal is to identify improved treatments for children and adolescents who have gained substantial weight on 2nd generation antipsychotic medications (SGAs). An improved treatment would: 1) provide adequate psychiatric symptom relief, 2) reduce SGA-induced weight gain, and 3) reduce SGA-induced insulin resistance and hyperlipidemia. Reductions in obesity, insulin resistance and hyperlipidemia are of great public health importance because they are factors strongly associated with type 2 diabetes and premature cardiovascular disease. The specific objective of the proposed study is to obtain data about the relative risks and benefits of 2 medication strategies for reducing SGA-associated weight gain and metabolic problems in youth, in comparison to control treatment. This study, which builds on extensive pilot data, will use a hybrid efficacy/effectiveness design to evaluate two competing medication approaches for the management of weight gain and metabolic side effects in youngsters on SGAs. The study is designed so that findings can be generalized to other clinical settings. We plan to enroll 240 participants, ages 8-17 years, who: 1) are currently treated with one of the three most commonly prescribed antipsychotics (risperidone, quetiapine or olanzapine), 2) have current BMI >85th percentile and have experienced substantial weight gain (>10%) during the past year while treated with their current SGA, 3) meet DSM-IV diagnostic criteria for a schizophrenia spectrum disorder or bipolar spectrum disorder, and 4) are psychiatrically stable. All participants will be randomized to one of three conditions: 1) continue current SGA (control group), 2) metformin + current SGA, or 3) staggered switch to aripiprazole with discontinuation of current SGA. The study will include a 3-week screening/baseline period and 24 weeks of treatment. The primary outcome variable is change in weight as reflected by change in BMI z-score. Secondary outcomes include: change in BMI percentile, change in weight as percent baseline weight, body fat mass, insulin sensitivity, lipids, prevalence of metabolic syndrome, and all cause treatment discontinuation. Individuals who experience continued excessive weight gain or psychiatric destabilization will be removed from the trial and treated as clinically indicated by the research team. The results of this project will inform future treatment for children and adolescents with major psychiatric disorders. The results are also likely to lead to treatments that improve their health and longevity.